The article shows that innovations are induced, since they become more profitable with the expansion of output. The amount of resources devoted to innovating activity, however, is in general not the optimal one because of the pressure of two opposing forces. On the one hand, competition between potential innovators tends to make this amount too large, on the other, the inability of innovators to capture all the benefits tends to make the amount too small. When all benefits are captured by the innovator either there is no economic growth due to innovations or else innovators are the sole beneficiaries from that growth. When benefits are diffused the innovation will always lead to economic growth, but only by sheer coincidence will it lead to maximum growth, which may be missed because the innovation is introduced either too early or too late. The rate of growth is always positive if the innovation is introduced too late. It may fall to zero with too-early introduction or even become negative if innovational activity is subsidized.